How do I clear JBoss' server.log file when JBoss is running? When I try to do
echo 1 > server.log
I get error msg that the file is being used by another program (JBoss). Is it possible to use a command-line tool (windows or linux(I do have CygWin)) or an application that I can write myself to clear that file?
P.S. I don't need that file to have 0kb, but I want it to have less than 100MB.
By default JBoss keeps the file locked, since it is writing log messages into it. It is locked as long as JBoss is running and I don't know of other way to release it than stopping JBoss itself.
To keep its size under control, you can modify your log configuration, which is by default in <server>/conf˛jboss-log4j.xml. You can specify the maximum size of a log file, and define what to do when that size is reached: roll over to a new file, truncate the existing one and start writing over it again, etc.
A basic example (not tested, so no guarantee that it works straight as it is):
<appender name="ROLL" class="org.apache.log4j.rolling.RollingFileAppender">
...
<param name="maxFileSize" value="100MB" />
...
</appender>
Moreover, with the maxBackupIndex parameter you may define the number of backup files (default is 1).
JBoss locks the file as long as the logging process is running.
If you enabled the JMX console you can stop the logging, delete / modify the log, and start the logging service again.
The url should look something like this (for log4j):
http://jboss.example.com:8080/jmx-console/HtmlAdaptor?action=inspectMBean&name=jboss.system%3Atype%3DLog4jService%2Cservice%3DLogging
I tested this with JBoss 5.
This solution should be scriptable as well.
Regarding your log file size problem: You should use a configuration approach instead of editing the log file manually.
Related
In a production environnement (for example) we often use INFO log level because with DEBUG log the size become quickly huge.
I know that's it's possible with SLJ4J or LOG4J to configure log rotation to split files etc.
But most of the time, we only need DEBUG logs when an exception appear, so we can have more information (HTTP Request logs for example).
I'm wondering if it's possible to keep only (for example) 1 minute of DEBUG log and save it if an exception appear. (Main log file is always on level INFO)
For example, when a HttpRequestException occurs, we have a file saved with 1 Minute of DEBUG log HttpRequestException-10-02-2015-10:28:50.log in a specific folder
So we can easily see all exceptions in this folder
/Exceptions/
- HttpRequestException-10-02-2015-10:28:50.log
- NullPointerException-10-02-2015-10:29:51.log
- IOException-10-02-2015-11:29:51.log
So... is it possible to do it with a tool like slf4j ?
And is DEBUG logging a huge performance problem on a production server?
Thanks
I found a solution.
I use cyclicBufferTracker to tell appender to use the last x lines once trigerred
And I can use the evaluator attribute to configure the trigger (default trigger is when an ERROR occurs)
Email is ok for me, but if someone want to put it in files instead of email I think you can create a custom Appender instead of SMTPAppender
Like on this page : http://jira.qos.ch/browse/LOGBACK-207
Here is the actual code : (You will need javax.activation and javax.mail if you use email implementation) and if you use gmail like me you will need to allow low security app in gmail settings : Gmail less secure apps
<appender name="EMAIL" class="ch.qos.logback.classic.net.SMTPAppender">
<smtpHost>smtp.gmail.com</smtpHost>
<smtpPort>587</smtpPort>
<asynchronousSending>false</asynchronousSending>
<STARTTLS>true</STARTTLS>
<username>GoogleUsername</username>
<password>mypassword</password>
<to>GoogleUsername#gmail.com</to>
<from>GoogleUsername#gmail.com</from>
<subject>Exception Alert</subject>
<layout class="ch.qos.logback.classic.PatternLayout">
<pattern>%date - %message%n</pattern>
</layout>
<cyclicBufferTracker class="ch.qos.logback.core.spi.CyclicBufferTracker">
<bufferSize>50</bufferSize>
</cyclicBufferTracker>
</appender>
Hope it will help
My current project taking Log4j for logging output but when it is deployed on websphere 7 it seems the logging file is always occupied by the server hence Log4J cannot close it and open a new file. Current log file can expand to 1.5 Gb if on DEBUG level. I went through a few online tutorials but didn't get a solid solution. So I open this thread for any opinion or experience on making log4J works on WS 7.
Current log4J is on 1.2.15, and the project is a plain POJO/JSP application without other containers(like Spring, I do have another project with same issue on Spring/tomcat, will open in another thread). And here is the log4j.properties:
log4j.rootCategory=INFO, ROL
#
# Rolling File Appender
#
log4j.appender.ROL=org.apache.log4j.RollingFileAppender
log4j.appender.ROL.Encoding=UTF-8
log4j.appender.ROL.File=C\:\\Logs\\AppOut.log
log4j.appender.ROL.MaxFileSize=10000KB
log4j.appender.ROL.MaxBackupIndex=10
log4j.appender.ROL.layout=org.apache.log4j.PatternLayout
log4j.appender.ROL.layout.ConversionPattern=%d -- %p -- %c -- %m%n
But this configuration just doesn't make log file roll over.
UPDATE:
It seems like each time when I try to remove the log file after stop the application(not the server), it is always saying "the file is opened by another application" which only refers to Websphere. I can guarantee that there is no other application is opening the log file. The only way to release lock on it is stop the application server.
Thanks in advance for any ideas.
we use was 7 (with portal server 6.1 on top). (We wrap our calls to the logger using the sl4j lbrary, but I don't think that is relevant to the problem here). can't really see any major difference, other than we don't set the root category, and we explicitly set the additivty to false.
Below is part of our config (showing just one of our appenders) which works just fine and rolls the logs every 10MB as expected. Perhaps it's something configured in WAS itself - I'm not a WAS admin expert and didn't install WAS myself but could ask one of our admins on Monday if you haven't solved it.
log4j.logger.com.xxx.protection=TRACE, A1
log4j.additivity.com.xxx.protection=false
# other appenders excluded
log4j.appender.A1=org.apache.log4j.RollingFileAppender
log4j.appender.A1.File=${LOG_ROOT}/applLogs/ui_and_business_logic.log
log4j.appender.A1.MaxFileSize=10MB
log4j.appender.A1.MaxBackupIndex=10
log4j.appender.A1.Append=true
log4j.appender.A1.layout=org.apache.log4j.PatternLayout
log4j.appender.A1.layout.ConversionPattern=%d %p %t %c MSG: %m%n
I take it that you've seen this article - http://www.ibm.com/developerworks/websphere/techjournal/0802_supauth/0802_supauth.html
i am having trouble with configuring Log4j correctly. I was expecting Log4j to rotate my catalina.out file at midnight when configuring it like the following..
log4j.properties:
log4j.rootLogger=INFO, CATALINA
# Define all the appenders
log4j.appender.CATALINA=org.apache.log4j.DailyRollingFileAppender
log4j.appender.CATALINA.File=/var/log/tomcat7/catalina.out
log4j.appender.CATALINA.Append=true
log4j.appender.CATALINA.Encoding=UTF-8
# Roll-over the log once per day
log4j.appender.CATALINA.DatePattern='.'yyyy-MM-dd-HH-mm'.log'
log4j.appender.CATALINA.layout = org.apache.log4j.PatternLayout
log4j.appender.CATALINA.layout.ConversionPattern =%d{HH:mm:ss} %5p [%t] - %m%n
After configuring I restarted Tomcat and
everything is written to:
/var/log/tomcat7/catalina.out
To test my configuration i changed the current date time to like 23:59:59:
#ls -l /var/log/tomcat7/
-rw-r--r-- 1 tomcat7 tomcat7 5840 4. May 00:00 catalina.out
As you can see, it didnt rotate at midnight... (?)
When restarting Tomcat it works perfectly fine:
#ls -l /var/log/tomcat7/
-rw-r--r-- 1 tomcat7 tomcat7 5840 4. May 13:37 catalina.out
-rw-r--r-- 1 tomcat7 root 2395 4. May 00:00 catalina.out.*CURRENTDATE*.log
Is it even possible to rotate my logfiles without restarting Tomcat?
Thanks in advance,
Marley
There are three solutions for this problem:
change default tomcat logging façade that writes to catalina.out to for example: slf4j, with all the benefits that comes with using it and log4j.
configure system cron to run logrotate of tomcat log files
change default logging class from ConsoleAppender to FileAppender.
Benefits of solutions:
very flexible as the slf4j offers many options especially with log4j, that you use anyway.
simple and doesn't require touching tomcat configuration.
simple change of configuration that disables console output
Disadvantages:
require additional libraries, affects all applications that tomcat is hosting, requires replacing default configuration with log4j.
cron+logrotate works only in linux; it might be not as simple in windows with scheduler. Requires extra scripting in windows environment.
provides only simple backup with date. Date pattern cannot be set. Does not compress rotated files.
Solution for First issue, is described here
Solution for Second issue is described here
Solution for Third issue is described here
You can as well combine solutions. For example use crontab to gzip files that where created by changing catalina.out to other name.
I would also suggest to leave tomcat so it logs to catalina.out, and configure your application to different file with log4j. This way logs from tomcat that are not immaterial won't spam your logs.
Is it even possible to rotate my logfiles without restarting Tomcat?
Yes, if you're willing to work for it.
Your log4j configuration will only end up fighting with the standard shell redirection that bin/catalina.sh uses to redirect stdout to logs/catalina.out. You can't simply use log4j configuration to change how System.out behaves.
If you want to rotate conf/catalina.out you will have to take some alternative measures depending on how to launch Tomcat:
If you use jsvc to launch Tomcat and you are using commons-daemon 1.0.4 or later, then you can send SIGUSR1 to the jsvc process to re-open the log files. That will allow you to move the existing log file to another file (which just changes its name and continues to log to the new filename) and then do 'kill SIGUSR1': the original filename will then be re-opened and new logs messages will go to it.
If you use bin/catalina.sh to launch Tomcat, you can modify it so that it no longer does redirection and instead pipes output to a rolling-logger process like Apache httpd's rotatelogs or chronolog.
I have a server I made in Java that needs to use a database, I chose HSQLDB.
So I have a lot of entries in my server like:
Logger.getLogger(getClass().getName()). severe or info ("Some important information");
When I run my server it goes to System.out which I think its the default configuration of java.util.logging?, so far its ok for me, and later I will make it go to a file ...
But, the problem is, when I start hsqldb it messes up with the default configuration and I can´t read my log entries on System.out anymore..
I already tried to change hsqldb.log_data=false, but it still messes up the default configuration.
Can someone help me??
I dont want to log hsqldb events, just my server ones.
Thanks
This issue was reported and fixed in the latest version 2.2.0 released today.
Basically, you set a system property hsqldb.reconfig_logging to the
string value false.
A system property is normally set with the -D option in the Java startup command for your application:
java -Dhsqldb.reconfig_logging=false ....
See below for details of the change:
http://sourceforge.net/tracker/?func=detail&aid=3195462&group_id=23316&atid=378131
In addition, when you use a fremework logger for your application, you should configure it directly to choose which levels of log to accept and which ones to ignore.
The hsqldb.applog setting does not affect framework logging and only controls the file log.
The hsqldb.log_data=false is for turning off internal data change logging and should not be used for normal databases. Its usage for bulk imports is explained in the Guide.
Try setting hsqldb.applog to 0, that shuts off application logging to the *.app.log file.
Start your server with a property pointing to the location of a dedicated properties file:
-Djava.util.logging.config.file=/location/of/your/hsqldblog.properties"
Which contains the following line to change Java logging for Hsqldb.
# Change hsqldb logging level
org.hsqldb.persist = WARNING
Side note, you can choose from the following levels:
SEVERE WARNING INFO CONFIG FINE FINER FINEST
I'm parsing gigantic Tomcat log files and I was wondering: when you stop, redeploy and then restart a Webapp, do the logs get automatically appended to the last debug.log.
More specifically: can you see in a unique debug.log file logs coming from two different deployments of a same .war?
So, for example, can you have logs from up to 11am from, say, version 1.0 of, say, example.war and then logs from 1pm coming from, say, version 1.1 of example.war in the same debug.log? Is this depending on the logger used and the way it is configured?
Tomcat will debug every error in the same log file, it doesnt matter if there is two different war files. Try using log4j (http://tomcat.apache.org/tomcat-5.5-doc/logging.html) it allows you to separate everything into different log files.